RESEARCH POSTER PRESENTATION DESIGN © 2012 www.PosterPresentations.com Fricatives are the rarest types of rhotics, reported for a few African and European languages [9] and as allophones in some Romance languages [4, 5, 8, 12]. Data from Nusu, a tonal Tibeto-Burman language of the Lolo/Yi/Ngwi subgroup spoken in China and Myanmar, demonstrate the presence of alveolar fricative rhotics in Southeast Asia. Nusu rhotics appear as onsets as well as the second consonant in initial clusters. Depending on environment and dialect, realizations include alveolar approximants [ɹ] as well as voiced and voiceless alveolar fricatives [ɹ̝ , ɹ̝̊]. In other studies on Nusu, the strident fricative realizations have also been transcribed as retroflex voiced sibilants [7, 13]. Our study challenges the tradition of interpreting strident alveolar fricatives as retroflex sibilants. Snatches of evidence from other Tibeto-Burman languages suggest that alveolar fricative rhotics are not limited to Nusu. These data provide further support for Whitley’s (2003) plea for an amended and more comprehensive treatment of rhotics, including voiced and voiceless alveolar fricatives. INTRODUCTION OBJECTIVES: The case for alveolar non‐sibilant fricaGves in Nusu ACOUSTIC EVIDENCE A. Alveolar approximant realizations Figure 1. Comparison of Nusu approximant variants in /pɹɔ̬/ ‘fly’: Alveolar [ɹ] in tense-voice Topya (left), palatalized in creaky-voice Myagu (right). REFERENCES B. Alveolar fricative realizations Figure 3. Change from approximation to frication in the second half of the initial consonant in [ɹɹ̝é] ‘to be swollen’ in Topya Nusu. Zoomed waveform to the right. Contact: elissa.ikeda@gmail.com ; sigrid_l@payap.ac.th 1. Present phonological evidence that the phones in question should be treated as rhotics based on phonotactics and shared phonetic features. 2. Provide acoustic data showing the range of phonetic realizations of Nusu /ɹ/, including voiced and voiceless alveolar fricatives. 3. Illustrate differences between alveolar non-sibilant fricatives and alveolar sibilants in Nusu. 4. Give evidence from other languages to demonstrate the challenges faced in transcribing non-sibilant fricative rhotics. Graduate Linguistics Department, Payap University, Chiang Mai, Thailand Elissa Ikeda & Sigrid Lew Alveolar Fricative Rhotics in Nusu METHODOLOGY PHONOLOGICAL EVIDENCE Recordings: 461 items by male speakers from Myanmar for the five Nusu varieties Myagu and Topya (2 speakers each), Wawa, Zileng, and Yotolo (1 speaker each). A subset of 105 items spoken by 4 additional Myagu speakers to investigate inter-speaker variation. The subset also was recorded with one additional speaker for Wawa, Zileng, Topya. Texts: Recorded readings of picture dictionary, New Testament segment, and story about Nusu origins were matched against the written texts to identify words containing a rhotic. They were identified if (a) a rhotic occurred in the text recordings of the corresponding word across Nusu varieties; (b) the word was spelled with <R> ; or (c) the word was transcribed with [ɹ] in one of the previous linguistic descriptions of Nusu. Programs used: Excel for lexical comparison across varieties; Phonology Assistant 3.3.3 (SIL 2011) for phonological analysis in Myagu and Topya; Speech Analyzer 2.7 (SIL 2005) for acoustic analysis of the Myagu and Topya data. Diachronic evidence: Syllable-initial consonant clusters are formed with the semi-vowels /j, w/, as seen in the examples [fjɑ̄ :] ‘tongue’ and [kʰwí] ‘dog’ from Myagu Nusu. Whereas the two former analyses of Nusu [7, 13] treat these semi-vowels as commencing parts of onglides, Bradley (1979) includes them with the liquids l and r as resonants in Proto- Loloish, positing possible frication for the rhotic. Despite the lack of rhotics in modern Loloish languages, reconstructed Proto-Loloish labial-rhotic and and velar- rhotic clusters *pr *br *kr *gr account for correspondences among the Loloish languages and the Burmish languages [2]. Nusu allows a rhotic, realized as alveolar approximants or fricatives and written with <R>, in clusters following labial and velar stops, nasals, and fricatives. These are the environments identified by the Chinese scholars for alveolar approximant rhotics in C2-position. Sonority: Nusu allows the rhotic /ɹ/ as well as semivowels /j, w/ as the second consonant in a cluster. Thus, rhotic alveolar approximant and fricative realizations are high in sonority. The rhotic can be preglottalized, as in /ɹɯ̄/ ‘retract’. Nusu allows only sonorants to be preglottalized, as seen in /nî/ ‘twist, /lɑ̂/ ‘turn sth. over’, /jû/ ‘person clf.’ Even though Nusu rhotics may be strident fricatives, they pattern like sonorants which is typical for rhotics [1, 16]. Cognate comparison across Nusus varieties: / ɹe / ‘swollen’ Myagu Topya (Zibankha) Topya (Ngwaphakha) Wawa Yotolo Zileng ɹ̝ǎ ɹ/ɹ̝/ʐé ɹˠé ɣɹǎm ɣɔǎ lua Figure 2. Comparing Myagu /j/ in [bjā] ‘to fly’ (left) and [ɹʲ] in the second syllable in [bɹʲábɹʲāxa] ‘to shine’ (right). F2 is higher for [j], accompanied by more turbulence in the higher spectrum. Figure 5. Comparison of spectrograms for initial voiced fricatives in Nusu. [ʐ] . Figure 4 (to the left). A more strident and partially sibilant-like realization of the second token of /ré/ by the same speaker. C. Sibilant realizations Former descrip/ons of Nusu include voiced sibilants /z, ʐ, ʑ/. Spectrograms of example words for Nusu frica/ves in these descrip/ons illustrate that the voiced retroflex sibilant /ʐ/ is not a separate phoneme, but a transcrip/on conven/on for the rho/c when it is realized as an alveolar frica/ve. A retroflex sibilant [ʐ] in medial position was mentioned in previous studies [7, 13]. The spectrogram shows more intensity than for [v, ɣ] but does not display the same level of high- frequency turbulence typical for sibilants and observed for [z, ʑ]. [1] Ballard, E, Starks, D. 2004. Liquids: Laterals and Rhotics or Much More? Proc. 2004 Australian Linguistic Society. [2] Bradley, D. 1979. Proto-Loloish. London and Malmø: Cuzon Press. [3] Bradley, D. 2012. The characteristics of the Burmish family of Tibeto-Burman. Language and Linguistics 13(1), 171–192. [4] Bradley, T.G. 2004. Gestural timing and rhotic variation in Spanish codas. In: Face, T.L (ed.) Laboratory approaches to Spanish phonology. Berlin: Walter de Gruyter. 195-220. [5] Colantoni, L., 2006. Increasing periodicity to reduce similarity: an acoustic account of deassibilation in rhotics. Díaz-Campos, M. (ed) Selected Proc. 2nd Laboratory Approaches to Spanish Phonetics and Phonology. 2006. Cascadilla Proceedings Project. Somerville: Cascadilla. 22-34. [6] Coupe, A.R. 2007. A grammar of Mongsen Ao. Berlin: Walter de Gruyter. [7] Fu, A. 1991. Nuyu [The Nu language]. In Dai, Q., Huang, B., Fu, A., Renzengwangmu & Liu Juhuang. (eds). ZangMianyu shiwu zhong [Fifteen Tibeto-Burman languages]. Beijing: Yanshan Chubanshe. 1252-1263. [8] Jesus, L.M.T., Shadle, C.H. 2005. Acoustic analysis of European Portuguese uvular [χ,ʁ] and voiceless tapped alveolar [ɾ̥ ] fricatives. JIPA. 35, 27-44. [9] Ladefoged, P., Maddieson, I. 1996. The sounds of the world's languages. Oxford: Blackwell. [10] Lubbe, M. 2005. Para Naga orthography statement. Unpublished document. SIL. [11] Maddieson, I. 1984. Patterns of sounds. Cambridge: Cambridge University Press. [12] Recasens, D. 2002. Weakening and strengthening in Romance revisited. Italian J. of Ling. 14, 327–373. [13]Sun H., Lu L. 1986. A brief overview of the language of the Nu nationality (Nusu language). Brief Overview of Minority Languages of China Series. Beijing: Nationalities Press. [14] Wells, J. C. 1982. Accents of English. Cambridge: Cambridge University Press. [15] Whitley, M.S. 2003. Rhotic representation: problems and proposals. JIPA 33, 81-86. [16] Wiese, R. 2001. The phonology of /r/. In Hall, T. A. (ed) Distinctive feature theory. 2001. Berlin: Walter de Gruyter. 335-368. Karbadian, Hopi, Eastern Armenian, Araucanian, Burushaski, Chukchi [11]), KiVunjo dialect of KiChaka, Czech, Edo [9], South African English [9, 14], Tibeto- Burman languages Mongsen Ao [6] and Para Naga [10]. CONCLUSION Nusu preserves the Proto Lolo-Burmese rhotic. It can be realized as a rhotacized vowel, a velar fricative, or as an alveolar approximant or non-sibilant fricative. Phonotactic patterning and the variety of place and manner of articulation indicate its status as a rhotic, even for its strident realizations. Transcriptions as retroflex sibilants might at least partially be caused by the lack of an individual alveolar fricative symbol. OTHER LANGUAGES WITH ALVEOLAR NON‐SIBILANT FRICATIVES 18 th International Congress of Phonetic Sciences, August 9-14, 2015, SECC Glasgow, UK